JP7351078B2 - Habit improvement devices, methods and programs - Google Patents

Description

The present invention relates to a habit improving device, method, and program.

Recently, there are many people (referred to as users) who are under pressure to improve their lifestyle habits based on the results of medical examinations and the like. For example, in response to a medical examination held by a workplace or local government organization, a user may be told that there is a high possibility of contracting a lifestyle-related disease in the future, or that the user has already contracted the disease. In such a case, the user may plan to change his daily life to improve his lifestyle, for example by seeking guidance from a specialist or using a specialized application. This plan will help you improve your daily routine by reviewing your current lifestyle habits in line with the symptoms of the disease, such as reducing the amount of food you eat, changing the type of food and drink you eat, and increasing your exercise time. Including designing your life.

For example, a pedometer is used to measure exercise time. A pedometer is a device that measures the number of steps by detecting vibrations of the human body, and is mainly used as a health device. Some pedometers have the function of calculating and displaying not only the number of steps, but also walking distance, walking speed, calories burned, etc. from the counted value of the number of steps and changes in the movement and posture of the human body. Devices having these functions can, for example, show each user the number of steps remaining at the current time with respect to the target value of the total number of steps per day, which is very convenient for users who are trying to improve their lifestyle habits.

For example, Patent Document 1 discloses a health support device that provides optimal advice for the user to continue walking comfortably by evaluating the user's step count in absolute number or amount of change. A health support device has been disclosed that aims to guide users to maintain and improve their health by providing more appropriate advice based on evaluation of step count levels and changes in step count. .

Patent Document 1 focuses on a method for evaluating the number of steps taken by a user, and determines the level of the number of steps taken in a recent certain period, and provides advice corresponding to the level, and also provides advice that matches the average number of steps up to one elapsed point in the past. It can be used to compare the average number of steps since the elapsed time and provide advice including the amount of change, and also to back up the data and update the advice text by sending the step count data to the server device via the information terminal. It is said that updates, etc., and changes in evaluation methods may be made in some cases.

Japanese Patent Application Publication No. 2004-121745

The conventional health support device described above evaluates the average step count level in a recent certain period based on the counted number of steps of the user, or compares the average number of steps up to one elapsed point in the past with the average number of steps after one elapsed point. By performing an evaluation that includes the amount of change, advice or information that is optimal for the user's step count situation is provided.

In this way, conventional devices provide advice to the user based on an evaluation of the user's step count based on an evaluation of the average step count level over a recent certain period, or an evaluation that includes the amount of change by comparing the average step count over different past periods. Or just providing information. For this reason, it is difficult to appropriately improve the user's lifestyle by not measuring activities in the user's lifestyle other than walking, and by not considering other factors other than walking that affect the user's lifestyle. There are some things that I can't do.

The present invention has been made in view of the above-mentioned circumstances, and in one aspect provides a habit improvement device, method, and program that guide a user to a suitable lifestyle by considering the user's activity level as well as factors other than the user's activity level. This is what I am trying to do.

The present disclosure employs the following configuration in order to solve the problems described above.

That is, the habit improvement device according to the first aspect of the present disclosure acquires the user's activity amount accumulated for each first period, and calculates the user's activity amount in a second period including the first period. a first calculation unit that calculates a distribution pattern indicating a change over time in the amount of activity; a classification unit that classifies the plurality of distribution patterns into one or more groups; and a goal that is a goal for the user based on factors other than the amount of activity. The device includes a second calculation unit that calculates patterns in association with the groups, and a presentation unit that presents advice information to the user based on the current distribution pattern and the target pattern.

In the above configuration, the first calculation unit calculates the distribution pattern indicating the temporal change in the user's activity amount over the second period, and integrates the activity amount for each first period included in the second period. , the distribution pattern of the activity amount is obtained in units of the second period divided into the first period. Typically, for example, the first period is one hour and the second period is one day, but any period may be used as long as the first period is included in the second period. The classification unit classifies the plurality of distribution patterns obtained thereby into one or more groups. The classification unit classifies the plurality of obtained distribution patterns using some method. This classification method includes, for example, clustering and class classification. Then, the second calculation unit acquires information regarding external factors brought about to the user (also referred to as external factors, external environment, and external environment), and based on this information, the user can achieve goals for each classified group. Calculate the target pattern that should be achieved.

The information regarding factors other than the amount of activity includes, for example, user attribute information and/or user external factor information. The user attribute information is information related to the user's attributes, and is information related to matters about the user itself, such as information indicating the user's gender, date of birth, and place of residence. Furthermore, unlike attribute information, the user's external factor information is information about matters other than the user, and is information about external factors brought about to the user. Note that details of the external factor information will be explained in the tenth aspect. Therefore, since the second calculation unit calculates the target pattern in association with the group, the target pattern is calculated in consideration of the distribution pattern based on the user's past activity amount. In other words, the second calculation unit calculates a desired target pattern for the user according to the user's attribute information and external factors brought about to the user, while considering the characteristics of the group.

The presentation unit obtains the distribution pattern of the amount of activity based on the current user's activities and the calculated target pattern, and presents advice information to the current user from these patterns to the user. The current distribution pattern indicates the distribution pattern from the beginning of the corresponding second period to the present, for example, the distribution pattern from the beginning of the day (e.g., waking up) to the present (e.g., 11 a.m. before lunch). shows. The presentation unit compares this current distribution pattern (i.e., history information of the amount of activity up to today) with the calculated target pattern, and determines, for example, if the amount of activity up to today is the cumulative amount of the goal pattern. If the amount of activity is less than the target pattern, advice information is presented so that the amount of activity can be consumed in the future, and in the opposite case, advice information is presented to make the amount of activity smaller than the target pattern. The presentation unit may present advice information to the user in consideration of external factors other than the amount of activity, such as future schedule, and attribute information.

Therefore, according to the present disclosure, the user can know advice for approaching ideal activity habits (e.g., walking habits) that also takes into account factors other than the amount of activity, such as external factors, and It becomes easier to take action to carry out. Further, according to the present disclosure, by classifying the amount of activity (e.g., number of steps) for each first period (e.g., one hour) using machine learning clustering, etc., a more detailed individual goal pattern for each first period can be created. In addition, by inputting personal external factors, it will be possible to present goals and encourage people to improve their health.
That is, according to the habit improvement device according to the first aspect of the present disclosure, patterns of activity amount are divided into groups instead of making a determination based only on the average value of activity amount as in the conventional technology. can discover that the current distribution pattern is abnormal compared to the usual pattern, inform the user that a different goal setting is required, and also present advice information for setting this goal. .

Further, according to the habit improvement device according to the first aspect of the present disclosure, by changing the first period and the second period, it is possible to obtain the period and advice timing of the desired distribution pattern and target pattern. For example, if (1st period, 2nd period) = (1 hour, 1 day), it is possible to know the daily distribution pattern that can change every hour, and provide advice information every hour. becomes possible. In addition, for example, if (1st period, 2nd period) = (1 day, 1 month), it is possible to obtain a monthly distribution pattern that can change every day. , second period)=(1 week, 1 year), the distribution pattern can be changed every week and a distribution pattern for one year can be obtained. By appropriately setting the first period and the second period in this manner, it is possible to obtain the distribution pattern and target pattern desired by the user. Furthermore, by adjusting these periods, it is possible to adjust the interval at which the user is provided with advice information, and the distribution pattern can be analyzed for both short, medium and long term, and the corresponding advice information can be adjusted. can be provided to the user.

Furthermore, the habit improvement device according to the first aspect of the present disclosure acquires the user's activity amount information and information regarding factors other than the activity amount (for example, attribute information and external factor information) and classifies the information into groups. Any device that can execute a program for generating target patterns and presenting advice information may be used, such as wearable devices (e.g., smartphones, wristwatch-type wearable terminals), activity meters, and stationary devices (e.g., , personal computer). Further, the habit improving device does not need to be worn by the user or not, as long as it can acquire activity amount information and the like. For example, the user may wear a detection device that can detect the amount of activity, and the habit improvement device may acquire activity amount information from this detection device and execute the above program.

In the habit improvement device according to the second aspect of the present disclosure, the classification unit is configured to generate a vector having the activity amount for each first period as a component, and one vector for each second period. an acquisition unit that corresponds to one distribution pattern and acquires distributions of the plurality of vectors from the plurality of distribution patterns calculated by the first calculation unit; and points in a space spanned by bases determined from the components; The vectors correspond to each other and include a clustering unit that clusters and classifies the plurality of points into one or more groups based on the positions of the plurality of points in the space.

In the above configuration, one vector for each second period, which has the activity amount for each first period as a component, corresponds to one distribution pattern in the second period. Furthermore, points in the space spanned by the basis determined from the components correspond to vectors. For example, the vector indicates a position vector, the starting point of the position vector corresponds to the origin where the amount of activity is zero in all the first periods included in a certain second period, and the ending point of the position vector corresponds to one distribution pattern. do. That is, in the acquisition unit, one vector corresponds to one distribution pattern in the second period for each second period, and the acquisition unit acquires the distribution of the plurality of vectors from the plurality of distribution patterns calculated by the first calculation unit. A space is prepared in which the number of dimensions is equal to or less than the number of first periods included in the second period, and each axis represents the amount of activity in the first period. For example, if the first period is one hour and the second period is one day, the maximum number of first periods is 24, so the space prepared in this case is at most a 24-dimensional space. Each axis corresponds to the user's activity amount for each hour from 0:00 to 24:00. In other words, the space in this example is an axis that shows the amount of activity from 0:00 to 1:00, an axis that shows the amount of activity from 1:00 to 2:00, and an axis that shows the amount of activity from 23:00 to 24:00. It becomes a 24-dimensional space including a maximum of 24 axes. Further, in this example, the space may be a space of 24 dimensions or less, for example, a 17-dimensional space divided by hour between 6:00 and 23:00, when the user is assumed to be active. Further, the acquisition unit calculates one point in space corresponding to the distribution pattern for each second period (for example, one day), and similarly calculates a plurality of points corresponding to the plurality of calculated distribution patterns. get. Therefore, the acquisition unit can acquire as many points as the number of distribution patterns in space. The clustering unit clusters the plurality of points into one or more groups based on the positions of the points distributed in space. As a result, the clustering unit can classify the distribution patterns into groups that are assumed to have similar distribution trends.

Therefore, according to the habit improvement device according to the second aspect of the present disclosure, the amount of activity for the second period or less for each first period is made to correspond to one point in space, and the distribution of this point is clustered to point By classifying into groups, it becomes possible to classify multiple distribution patterns.

In the habit improvement device according to the third aspect of the present disclosure, the clustering unit compares distances between a certain point and another point in the space, and groups two points with the smallest distance as belonging to the same group, Compare the distances between a point belonging to a group and multiple points belonging to other groups, and if the minimum distance among these distances is less than or equal to a threshold, the points that have the minimum distance are grouped together. are grouped as the same group.

In the above configuration, the clustering unit groups a plurality of points by finding distances between points distributed in a (generally multidimensional) space and comparing the distances. Although it is desirable that the distance in space used in this embodiment satisfies the axiom of mathematical distance, the axiom of mathematical distance is not a necessary and sufficient condition for the distance in this embodiment. More specifically, the distance axiom is a necessary and sufficient condition for mathematical distance, but is not a necessary condition for the concept of distance in embodiments of the present application. Moreover, the distance here is an expanded concept of distance in one, two, or three-dimensional space, and can have various definitions. Typical examples of these distance definitions are Euclidean distance, Citiblock distance, Minkowski distance, Mahalanobis distance, and cosine similarity (a concept opposite to distance). Distance is also called dissimilarity, and is a function that measures the size of comparison targets in data or clusters. Although any distance, dissimilarity, and/or similarity may be used in this disclosure, in this description, for convenience, distance is used to encompass all concepts discussed herein.

In addition, although the habit improvement device according to the third aspect shows a typical specific example executed by the clustering unit, various other methods (agglomerative (or segmented) hierarchical clustering, optimal clustering, etc.). For example, the agglomerative hierarchical clustering method may be performed using the following procedure. First, it starts with the assumption that each point forms an isolated group, calculates the distance between group pairs for all these groups, and finds the group pair that has the minimum distance. Then, these group pairs are merged to generate a new group, and the distance between this new group and other groups is calculated. Next, among the group pairs including this new group, a group pair with the minimum distance is found, and these group pairs are merged to generate a new group. After that, repeat the merging of groups and recalculation of the distance in the same way, proceeding until the entire point becomes one group, return to the appropriate number of groups or the group at the time of merging, etc., and increase the number of groups. Determine. The clustering unit can advance grouping based on distance (dissimilarity) and divide a plurality of points into groups according to the procedure described above. Note that when determining the number of groups, when the desired number of groups is reached before all the points become one group, that group may be fixed.

The habit improvement device according to the fourth aspect of the present disclosure further includes a pattern generation unit that generates one of the distribution patterns belonging to the group for each group as a typical pattern that is a typical pattern of the group. Be prepared.

In the above configuration, the pattern generation unit generates, for each group, a pattern that is typical of that group. The habit improvement device according to the fourth aspect generates one of the distribution patterns belonging to a group as a typical pattern. Therefore, the typical pattern generated by the pattern generation section can be used when handling groups. Since groups can be handled as patterns, when the second calculation unit calculates target patterns in association with groups, it can perform calculations based on typical patterns.

The habit improvement device according to the fifth aspect of the present disclosure further includes a pattern generation unit that generates a distribution pattern corresponding to a point determined by the distribution of points belonging to the group as a typical pattern that is a typical pattern of the group. Be prepared.

In the above configuration, the pattern generation unit generates a pattern that is typical for each group based on the distribution of points in space. In the habit improvement device according to the fifth aspect, a distribution pattern corresponding to a point determined by a distribution within a group of points indicating a distribution pattern belonging to a group is generated as a typical pattern. Examples of points determined by the distribution within a group of points indicating a distribution pattern belonging to the group include the center of gravity determined by a plurality of points within the group, and the center of the space in which the group is distributed. Therefore, the typical pattern generated by the pattern generation section can be used when handling groups. Since groups can be handled as patterns, when the second calculation unit calculates target patterns in association with groups, it can perform calculations based on typical patterns.

In the habit improvement device according to the sixth aspect of the present disclosure, the classification unit classifies into the groups based on whether the amount of activity in a specific time period of the distribution pattern is equal to or higher than a threshold value.

In the above configuration, the distribution pattern of the activity amount is classified into groups based on whether the amount of activity in the specific time period of the distribution pattern is large (that is, whether it is equal to or greater than a threshold value). In the method according to the sixth aspect, for example, if the amount of activity is large after midnight (for example, until 2 o'clock), if the amount of activity is large before 6 o'clock in the morning (for example, after 4 o'clock), and/or during the daytime (for example, from 9:00 to 15:00), if the amount of activity is small (that is, smaller than a certain threshold), the classification unit classifies the distribution patterns into corresponding groups. The characteristic viewpoint of the habit improvement device according to the sixth aspect is that, like the habit improvement device according to the fifth aspect, the conditions that define the group (for example, whether it is a typical pattern, whether it is a typical pattern, The difference is that the amount of activity (whether the amount of activity is above the threshold value) is set in advance, and a setter such as a user can set the characteristics of the group in advance. In the habit improvement device according to the sixth aspect, since the activity amount in a specific time period can be set as a threshold in advance, the activity amount in a specific time period can be grouped by the threshold value. It becomes possible to assign meaning in advance.

In the habit improvement device according to the seventh aspect of the present disclosure, the second calculation unit refers to schedule information regarding the user's schedule and location information of the user, and determines whether the user is scheduled in the schedule. It is determined whether the target pattern has deviated from the position, and if it is determined that the target pattern has deviated from the position, the target pattern is changed according to the deviated position, and the presentation unit displays the current distribution pattern and the changed target pattern. The advice information is presented based on the determined target pattern.

In the above configuration, the second calculation unit refers to the user's schedule information and the user's location information to determine whether the user's current location deviates from the location scheduled in the schedule. For example, the second calculation unit determines whether the user's current location is more than a preset distance away from the location scheduled in the schedule information. If it is determined that the user is further away than this distance, the second calculation unit determines that the user has deviated from the expected position. If it is determined that the user has deviated from the planned position, the second calculation unit calculates the amount according to the deviated position (the location or the building existing at the location linked to the location). to change the target pattern. For example, if the deviating position is a gym, it is considered an opportunity to increase the amount of activity, and a target pattern to increase the amount of activity is generated, whereas if the deviating position is a restaurant, increases the amount of activity more than usual, assuming the time after eating at a restaurant or taking a break. The second calculation unit estimates the time of leaving the restaurant by estimating the meal content from the time of entering the restaurant, for example. The presentation unit can present appropriate advice information to the user by presenting the advice information to the user based on the current distribution pattern and the changed target pattern.

In the habit improvement device according to the eighth aspect of the present disclosure, the presentation unit includes a calculation unit that calculates a difference between the current distribution pattern and a target pattern corresponding to the current distribution pattern for each first period. , an advice generation unit that generates the advice information based on the difference.

In the above configuration, the presentation unit calculates the difference between the current distribution pattern and the target pattern corresponding to this distribution pattern, and calculates the difference including the sign, which is the minimum unit of the distribution pattern in which the amount of activity is accumulated. Based on this difference, how much larger or how much smaller the current distribution pattern is than the target pattern is calculated for each first period. The presentation unit then generates advice information to be presented to the user based on the difference between the first periods. The advice information is a comment for improving the user's lifestyle habits, and is, for example, a comment for urging the user to increase the amount of activity in the future because the amount of activity according to the distribution pattern up to now is less than the target pattern. Note that the presentation unit may calculate the amount of activity and the time period based on information other than the amount of activity (for example, user's external factor information). In this case, for example, the presentation unit may indicate that it is a time when the user is scheduled to be active according to the schedule, and if it is a time when the user is scheduled to be active, it is possible to be active outside during this time according to the weather forecast. The amount of activity and the time period are calculated based on whether or not the activity level is the same, and corresponding advice information is generated.

In the habit improvement device according to the ninth aspect of the present disclosure, the first calculation unit uses at least one of the amount of energy consumed by the user and the number of steps taken by the user as the amount of activity.

In the above configuration, the first calculation unit calculates the amount of activity using the amount of energy consumed by the user and/or the number of steps taken by the user. For this reason, the first calculation unit can measure energy consumption using, for example, an activity meter, and can also measure the number of steps using a pedometer. The activity meter uses, for example, an acceleration sensor (three-dimensional acceleration sensor) and an atmospheric pressure sensor to detect changes in the movement and posture of the human body, identify various activities, and calculate the activity amount. Activities are related to human movement, and indicate physical activities such as walking, jogging, cleaning, and doing laundry, for example. Walking can also be identified as normal speed, slow speed, brisk walking speed, etc., for example. Energy consumption indicates the total amount of energy consumed by the user during a certain period (for example, one day).

In the habit improvement device according to the tenth aspect of the present disclosure, the second calculation unit may include an external factor related to an external factor that is a factor brought about to the user by matters other than the user, as a factor other than the activity amount. Based on the information, the target pattern is calculated in association with the group.

In the above configuration, the user's external factor information is information about matters other than the user, unlike the user's attribute information, and is information about external factors brought to the user, such as information about the user's current location. weather forecast information corresponding to the user's location information, and information regarding the user's schedule. External factors here refer to influences brought about by the environment outside the user, that is, the environment surrounding the user. Therefore, the second calculation unit can calculate a target pattern associated with a group that is desirable for the user according to external factors brought about to the user, while taking into consideration the characteristics of the group.

In the habit improvement device according to the eleventh aspect, the second calculation unit includes, as the external factor information, location information of the user, weather forecast information based on the location information, and schedule information regarding the user's schedule. Use at least one of these.

In the above configuration, external factor information used in the second calculation section is defined, and the external factor information is information used to calculate a target pattern that is a goal for the user by associating it with the group. location information indicating the user's current location, weather forecast information based on location information (e.g., weather conditions for each hour in the future), and/or schedule information for the user (e.g., weather conditions for each hour in the future). (planned). These external factor information enable the second calculation unit to calculate accurate target patterns in association with groups. In addition to location information, it may also include location history information that records the history of this location, and may also record or predict the ease of doing activities depending on a specific location and include this in the external factor information. Good too. The external factor information may be any external factor brought to the user, and is not limited to the items described here.

In the habit improvement device according to the twelfth aspect, the presentation unit compares the current distribution pattern with a target pattern corresponding to the current distribution pattern and presents the advice information to the user.

In the above configuration, for example, the presentation unit compares the current distribution pattern with the corresponding target pattern and presents the result to the user, allowing the user to know which part of the pattern has reached the target and which other parts have reached the target. You will be able to know what kind of activities you should do in the future and when. For example, if the distribution pattern is such that the amount of activity is recorded as a time history, the user can compare the target and actual amount of activity for each unit of time in which the amount of activity is measured, and the user can use it to achieve the goal. It becomes possible to easily recognize the amount of activity required for

In the habit improvement device according to yet another aspect, the classification unit classifies the habits into the groups based on a plurality of preset typical patterns, which are typical patterns of the groups.

In the above configuration, the classification unit classifies the plurality of distribution patterns into groups using a preset typical pattern indicating the characteristics of each group. A characteristic aspect of the habit improvement device according to this aspect is that typical patterns are set in advance. In this aspect, the typical pattern can be set in advance by a setter such as a user. Since typical patterns can be set in advance, it is possible to assign meaning to typical patterns in advance. In the habit improvement device according to the second aspect, similar distribution patterns are collected to form one group, and the characteristics of the group are given meaning from the formed group.On the other hand, in the habit improvement device according to this aspect, the group Before forming the group, the characteristics of the group (in this case, a typical pattern) are first determined, groups are set based on this characteristic, and distribution patterns are classified based on this group.

Also, the method of the second aspect and the method of this aspect may be used. For example, the distribution patterns are grouped using the method of the second aspect, a typical pattern of the group that the user considers to be useful is set for each group based on the group, and then the distribution patterns are grouped again using the method of this aspect based on this typical pattern. may be converted into As a result, it is possible to set typical patterns using this method based on the actual distribution of distribution patterns using the second method. becomes possible.

In the habit improvement device according to still another aspect, the second calculation unit includes attribute information regarding attributes of the user as a factor other than the amount of activity, and the attribute information includes physical information and residence information of the user. , occupation, place of work, hobbies, and eating and drinking preferences.

In the above configuration, the attribute information used in the second calculation unit is defined, and the attribute information is information used when calculating the user's target pattern, for example, when measuring the user's activity amount. There is physical information necessary for personal information, information on residence, occupation, and place of work to correlate with external factors, information on hobbies related to activities, and information on eating and drinking preferences used when eating out. Further, the attribute information may be used by the presentation unit to present appropriate advice information.

In the habit improvement device according to still another aspect, the classification unit generates a vector having the activity amount for each first period as a component, and one vector for each second period corresponds to one distribution pattern in the second period. , one point corresponding to one of the distribution patterns in space is calculated for each second period including the first period, and a plurality of the vectors are calculated from the plurality of distribution patterns calculated by the first calculation unit. an acquisition unit that acquires the distribution of the plurality of points, the points in the space spanned by the basis determined from the component and the vector correspond to each other, and the plurality of points are determined based on the positions of the plurality of points in the space. a clustering unit that clusters and classifies into one or more groups, and the space has each axis corresponding to a first period of an amount indicated by a certain index of the activity amount, and a clustering unit that clusters and classifies the activity amount into one or more groups, It is a space consisting of each axis corresponding to the first period of quantity indicated by the index.

In the above configuration, the acquisition unit corresponds to each axis corresponding to the first period of the amount indicated by a certain indicator of the activity amount and the first period of the amount indicated by another indicator of the activity amount. Since there are axes for each axis, the maximum number of dimensions is twice the number of first periods included in the second period. For example, one indicator of the amount of activity is the number of steps, and another indicator of the amount of activity is energy consumption. In this example, if the first period is one hour and the second period is one day, the space will be a maximum of 48 dimensions. A space is prepared in which each axis indicates one index and another index of the amount of activity in the first period. For example, if the first period is one hour and the second period is one day, the maximum number of first periods is 24, so the space prepared in this case has two types of indicators. (In other words, there are two types of "certain index" and "another index" mentioned above), resulting in a maximum of 48-dimensional space, and for each index, each axis corresponds to the user's activity amount for each hour from 0:00 to 24:00. do. In other words, the space in this example has 2 axes showing the amount of activity from 0:00 to 1:00, 2 axes showing the amount of activity from 1:00 to 2:00, etc., the amount of activity from 23:00 to 24:00. It becomes a 48-dimensional space including a maximum of 2 x 24 axes, with two axes representing . Further, the space in this example may be a space of 48 dimensions or less, for example, it may be a 34-dimensional space divided by hour from 6:00 to 23:00, when the user is assumed to be active.

Further, the acquisition unit calculates one point in space corresponding to the distribution pattern for each second period (for example, one day), and similarly calculates a plurality of points corresponding to the plurality of calculated distribution patterns. get. Therefore, the acquisition unit can acquire as many points as the number of distribution patterns in space. The clustering unit clusters the plurality of points into one or more groups based on the positions of the points distributed in space. As a result, the clustering unit can classify the distribution patterns into groups that are assumed to have similar distribution trends. Further, according to this habit improvement device, by focusing on different types of activity amounts, it is possible to make the characteristics of a group more prominent.

Further, the type of activity amount corresponding to the spatial axis can be selected from other types, and does not need to be a specific type. In addition to those listed here, the types of activity amount include, for example, activity energy, basal metabolism, meal-induced body heat production, number of fast steps, and number of stairs climbed.

According to one aspect of the present invention, it is possible to provide a habit improvement device, method, and program that guide a user to a suitable lifestyle by taking into consideration not only the user's activity level but also factors other than the user's activity level.

1 is a diagram showing an overview of a system including a habit improvement device, an activity meter, a wristwatch-type wearable terminal, and a server connected via a network according to an embodiment. FIG. 1 is a diagram schematically illustrating an example of a hardware configuration of a habit improvement device according to an embodiment. 1 is a diagram illustrating an example of a part of a software configuration of a habit improvement device according to an embodiment; FIG. 1 is a one-day time chart showing the operation of the habit improvement device and the user's behavior according to the embodiment. 1 is a flowchart schematically illustrating an example of a processing procedure regarding the habit improvement device according to the embodiment. 6 is a flowchart schematically illustrating an example of the processing procedure of step S502 in FIG. 5. FIG. 6A is a table showing an example of activity amount information obtained by the processing procedure of FIG. 6A. 6 is a flowchart schematically illustrating an example of the processing procedure of step S503 in FIG. 5. FIG. 7A is a flowchart schematically illustrating a detailed example of the processing procedure of steps S701 and S702 in FIG. 7A. FIG. FIG. 7B is a diagram showing an example of distribution patterns obtained by the processing procedure of FIG. 7B and groups into which they are classified. 6 is a flowchart schematically illustrating an example of the processing procedure of step S504 in FIG. 5. FIG. 9 is a flowchart schematically illustrating an example of the processing procedure of step S802 in FIG. 8A. 8B is a diagram showing an example of a typical pattern obtained in step S812 or a basic pattern obtained in step S817, and a target pattern obtained in step S818. FIG. The figure which shows an example of generating a target pattern from a typical pattern or a basic pattern according to a schedule. FIG. 8D is a diagram illustrating an example of confirming user behavior based on the schedule shown in FIG. 8D based on location information. 6 is a flowchart schematically illustrating an example of the processing procedure of step S505 in FIG. 5. FIG. 9A is a flowchart schematically illustrating an example of the processing procedure of step S901 in FIG. 9A. 9A is a diagram schematically illustrating presentation of advice information obtained in step S902 of FIG. 9A. FIG.

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment (hereinafter also referred to as "this embodiment") according to one aspect of the present invention will be described below based on the drawings. Note that in the following embodiments, portions with the same numbers perform similar operations, and redundant explanation will be omitted.

[overview]
First, the outline of the habit improving device of the present invention will be explained using FIG. FIG. 1 schematically illustrates a habit improvement device 100, an activity meter 110, a wristwatch-type wearable terminal 120, a server 130, a network 140, and a GPS satellite 150 according to an example of the outline. The activity meter 110 measures the amount of activity of the user by being carried by (or worn by) the user, and detects changes in the movement or posture of the human body using, for example, an acceleration sensor and/or a barometric pressure sensor. Identify the user's various activities and calculate the amount of activity.

The amount of activity is the amount of energy consumed, which is the total amount of energy consumed by the user during a certain period (for example, one day). Energy consumption corresponds to the total energy consumed by the user in a certain period of time, and is expressed as, for example, the sum of active energy, basal metabolism, and meal-induced body heat production. Activity energy is the energy consumed by the user's physical activities and exercise, and basal metabolism is the energy necessary to maintain life (maintaining body temperature, moving the heart, etc.) and is consumed even when the user does not move his body. It is an energy that Furthermore, meal-induced body heat production is the energy consumed by activities such as digestion and absorption of food after a meal. Activity energy is calculated, for example, by the product of MetS value, body weight, exercise time, and a constant. The MET value is a value indicating how many times the amount of energy consumed during exercise corresponds to that at rest, and is expressed in units of METs. Further, the MetS value is determined by subdividing each exercise and activity.

However, here, the amount of activity may be defined as any one or more of active energy, basal metabolism, and meal-induced body heat production, for example, as only active energy, only active energy, and only basal metabolism. Good too. Furthermore, since it is a good approximation to simply use the number of steps as the amount of activity, the amount of activity may also be the number of steps.

The activity meter 110 measures any one or more of energy consumption or active energy, basal metabolism, and meal-induced body heat production, and includes, for example, the number of steps, the number of brisk steps, the number of steps taken while walking slowly, and the number of stairs. The number of steps uphill may be measured, or only the number of steps may be measured. The wristwatch-type wearable terminal 120 can also measure the amount of activity similar to the amount of activity meter 110.

Furthermore, the habit improvement device 100 integrates the user's distribution pattern for each first period (for example, one hour) based on the amount of activity acquired from the activity meter 110 and/or the wristwatch-type wearable terminal 120, and calculates the distribution pattern of the user for each first period (for example, one hour). A distribution pattern over a large second period (for example, one day) is calculated. The habit improvement device 100 then classifies the calculated distribution patterns into a plurality of groups. Methods for classifying into groups include, for example, clustering, class classification, etc., but any method may be used as long as it is possible to group. After grouping, the habit improvement device 100 calculates a target pattern that is the user's goal for each group based on the user's attribute information and external factor information for the user. Then, the habit improvement device 100 compares the target pattern and the current distribution pattern, generates advice information for the user, and presents it to the user.

Note that the distribution pattern is expressed as a pattern distributed in a two-dimensional space in which the horizontal axis represents time and the vertical axis represents the amount of activity. For example, for each first period, the total amount of activity within this period is calculated. The indicated frequencies are displayed in a two-dimensional space for the second period in association with time. As a specific example, the distribution pattern is a bar graph that represents the sum of the amount of activity measured every hour from the time the user wakes up until before going to bed, and a bar graph for the period during which the user is awake. It is a list of That is, the distribution pattern may be any pattern that expresses the temporal change in the total amount of activity for each period.

For example, a program executed by the habit improvement device 100 for calculating distribution patterns, grouping the distribution patterns, calculating target patterns for each group, and presenting advice information is executed via the network 140. is obtained from the server 130 that provides the. The user's attribute information may also be acquired via the network 140 from the server 130 that has the corresponding attribute information, or the user may input the attribute information through the input unit of the habit improvement device 100. Note that this program may of course be stored in the habit improvement device 100 in advance. Further, the external factor information is information related to external factors that are factors brought about to the user, other than information related to the user himself/herself, and includes, for example, location information, weather information, user schedule information, and the like. The habit improvement device 100 also acquires this external factor information from the server 130 via the network 140. Note that the user's location information is calculated when the habit improvement device 100 receives a signal from the GPS satellite 150, but the habit improvement device 100 receives information from a base station and/or a wireless LAN access point and uses this information. The position information may be corrected accordingly.

Note that if the habit improvement device 100 can independently detect the amount of activity detected by the activity meter 110 and/or the wristwatch-type wearable terminal 120, the habit improvement device 100 does not use the activity meter 110 or the like. The amount of activity detected by 100 may be used. In this case, the habit improvement device 100 will include a device portion for calculating the activity amount of the activity meter 110 and/or the wristwatch-type wearable terminal 120. For example, since the habit improvement device 100 is a wearable terminal device (for example, a smartphone), it is normally equipped with an acceleration sensor and a pressure sensor, and can measure the number of steps and the number of floors. It becomes possible to calculate the amount of activity based on the floor number.

As described above, according to the habit improvement device of the present embodiment, the user can receive advice for approaching ideal activity habits (e.g., walking habits) that also take into account external factors, and can implement the activity habits. It is expected that users will be able to easily take action to do so. In addition, according to this habit improvement device, by classifying the amount of activity for each first period using clustering etc., it becomes possible to calculate a more detailed individual goal pattern for each first period. By considering external factors, it becomes possible to present goals and encourage people to improve their health. Furthermore, the habit improvement device of this embodiment does not make a determination based only on the average value of the amount of activity as in the conventional technology, but by dividing the pattern of the amount of activity into groups, the current distribution pattern is It is possible to detect an abnormality by comparing the pattern with the above pattern, inform the user that a different goal setting is required, and also present advice information for this goal setting.

[Configuration example]
(Hardware configuration)
<Habit improvement device>
Next, an example of the hardware configuration of the habit improvement device 100 according to this embodiment will be described using FIG. 2.
As shown in FIG. 2, the habit improvement device 100 according to the present embodiment includes a communication interface 201, a storage section 202, an input device 203, an output device 204, a control section 205, a timekeeping device 206, a power supply section 207, and a GPS reception section 208. , and an external interface 209 are electrically connected to the computer. The habit improvement device 100 according to this embodiment corresponds to the "habit improvement device" of the present invention. In addition, in FIG. 2, the communication interface and external interface are described as "communication I/F" and "external I/F", respectively.

The communication interface 201 is, for example, a short-range wireless communication (e.g., Bluetooth (registered trademark)) module, a wired LAN (Local Area Network) module, a wireless LAN module, etc., and is used for wired or wireless communication via a network. It is an interface. The communication interface 201 is an interface for connecting the habit improvement device 100 to an external device (for example, a computer, a communication device on a network). The communication interface 201 is controlled by the control unit 205 and is for receiving activity amount information from the activity meter 110 and/or wristwatch-type wearable terminal 120, and also receives external factor information and user information via the network 140. This is for downloading attribute information and/or a program to be executed by the habit improvement device 100 from the server 130 or the like.

Communication via this network may be wireless or wired. Note that the communication interface 201 may be able to transmit information to an external device via a network. The network may be an internetwork including the Internet, another type of network such as a hospital LAN, or one-to-one communication using a USB (Universal Serial Bus) cable or the like. good. Communication interface 201 may include a micro USB connector.

The storage unit 202 stores information such as programs electrically, magnetically, optically, mechanically, or chemically so that computers, other devices, machines, etc. can read the recorded information such as programs. It is a medium that accumulates by action. The storage unit 202 is, for example, an auxiliary storage device such as a hard disk drive or a solid state drive, and stores activity amount information acquired from the activity meter 110 and/or wristwatch-type wearable terminal 120 and pattern information (calculated by the control unit 205). For example, distribution pattern information and target pattern information), user attribute information input from the input device 203 or acquired via the network 140, user external factor information acquired via the network 140, And/or an execution program for lifestyle improvement presentation, which is executed by the control unit 205 and presents information for lifestyle improvement to the user based on the acquired activity amount information, is stored. Regarding the activity amount information acquired by the activity meter 110 and the like, all activity information from a certain time in the past may be stored in the storage unit 202. Furthermore, the storage unit 202 may also store group information regarding classified groups determined by the distribution pattern. Further, advice information may be stored in association with the difference pattern calculated by the execution program. This advice information and the difference pattern may be associated with each other to form a table.

The input device 203 is a device that receives input, and is, for example, a touch panel, physical button, mouse, keyboard, or the like. The output device 204 is a device that performs output, and outputs information by display, audio, etc., and is, for example, a display, a speaker, or the like.

The control unit 205 includes a CPU (Central Processing Unit), a RAM (Random Access Memory), a ROM (Read Only Memory), and the like, and controls each component according to information processing. An execution program that presents lifestyle improvements to the user based on the acquired activity amount information is stored in the storage unit 202, and the control unit 205 calls the execution program from the storage unit 202 and executes the process. In addition, the control unit 205 performs distance measurement calculations based on GPS signals received via a GPS reception unit 208 (described later), and obtains current position information of the habit improvement device 100, that is, when the activity meter 110 or the wristwatch-type wearable terminal 120 is worn. Calculate the location of the person being measured (user).

The clock device 206 is a device that measures time and can measure date and time. For example, the clock device 206 is a clock including a calendar, and passes information on the current year, month, and/or date and time to the control unit 205.

The power supply unit 207 may be anything that can supply power, such as a rechargeable secondary battery or an AC power source that can be obtained from a normal outlet. The power supply unit 207 supplies power to each element installed in the main body of the habit improvement device 100. The power supply unit 207 supplies power to, for example, the communication interface 201, the storage unit 202, the input device 203, the output device 204, the control unit 205, the clock device 206, the GPS reception unit 208, and the external interface 209.

GPS receiving section 208 receives respective GPS signals transmitted from a plurality of GPS satellites, and outputs the received GPS signals to control section 205. The GPS receiving unit 208 may be included in the activity meter 110 and/or the wristwatch-type wearable terminal 120 instead of the habit improvement device 100, and these acquire the GPS signal and send it to the habit improvement device 100 via the communication interface 201. may receive the GPS signal, and the control unit 205 may calculate the position information. Alternatively, the activity meter 110 and/or the wristwatch-type wearable terminal 120 may receive a GPS signal, calculate position information, and transmit the position information to the habit improvement device 100.

The external interface 209 serves as an intermediary between the main body of the habit improvement device 100 and the outside, and is, for example, a USB port, etc., and is an interface for connecting to an external device (for example, a memory, a communication device). . The external interface 209 is, for example, an interface for connecting to an external device such as an acceleration sensor, a gyro sensor, and/or a geomagnetic sensor.

(Software configuration)
<Habit improvement device 100 lifestyle improvement suggestion presentation>
Next, an example of the software configuration of the habit improvement device 100 according to this embodiment will be described using FIG. 3. FIG. 3 shows how the control unit 205 of the habit improvement device 100 classifies the distribution patterns calculated based on the user's activity level into groups, calculates the target pattern for each group, and calculates the target pattern based on the target pattern and the current distribution pattern. This figure shows a software configuration for executing a program for presenting advice information to users.

When executing the necessary program, the control unit 205 of the habit improvement device 100 executes an execution program that presents advice information regarding lifestyle improvement to the user based on the acquired activity amount information stored in the storage unit 202. Expand to RAM. The control unit 205 then uses the CPU to interpret and execute an execution program developed in the RAM that presents advice information regarding lifestyle improvement to the user, thereby controlling each component. As shown in FIG. 3, the habit improvement device 100 according to the present embodiment includes an activity amount acquisition section 301, a distribution pattern calculation section 302, a group classification section 303, an external factor information acquisition section 304, and an attribute information acquisition section. section 305, target pattern calculation section 306, and advice presentation section 307.

The activity amount acquisition unit 301 acquires activity amount information of a desired user from an external activity meter 110 or the like via the communication interface 201 . The activity amount acquisition unit 301 acquires activity amount information acquired by an external activity amount sensor via the communication interface 201 at regular intervals. The fixed period is, for example, every hour, but the activity amount acquisition unit 301 may acquire the activity amount information at a shorter period, such as every minute. Further, the external activity sensor is an activity meter 110, a wristwatch-type wearable terminal 120, etc., and the habit improvement device 100 is equipped with these devices, and the activity amount acquisition unit 301 acquires activity amount information from these devices. You may.

The distribution pattern calculation unit 302 integrates (accumulates) the amount of activity acquired by the activity amount acquisition unit 301 for each first period, and calculates a distribution pattern in which the first period is combined into a second period. Hereinafter, the case where the first period is one hour and the second period is one day will be explained as a main example. When the amount of activity is the number of steps, the amount of activity acquisition unit 301 calculates the sum of the number of steps walked by the user during that time period every hour. The distribution pattern calculation unit 302 calculates the number of steps for 24 hours, for example, and obtains a bar graph with time on the horizontal axis and number of steps on the vertical axis. In this example, the distribution pattern calculation unit 302 obtains a bar graph for 24 steps (including 0 steps) corresponding to the time period as the distribution pattern for one day. The distribution pattern calculation unit 302 calculates distribution patterns for a plurality of second periods and obtains a plurality of distribution patterns. The distribution pattern calculation unit 302 calculates the distribution pattern on a daily basis, for example, but if past activity amount information is in the storage unit 202 at the time when the habit improvement device 100 is started, the distribution pattern calculation unit 302 calculates the distribution pattern from the past up to the present. The distribution pattern may be calculated, or the distribution pattern may be calculated for each first period.

The group classification unit 303 classifies the plurality of distribution patterns calculated by the distribution pattern calculation unit 302 into one or more groups. The group classification unit 303 classifies groups such that different groups have different characteristics. The group classification unit 303 uses clustering, class classification, etc. as a classification method, and does not particularly limit the classification method. A more specific example will be described later with reference to FIGS. 7A, 7B, and 7C.

When the group classification unit 303 classifies groups by clustering, for example, a 24-dimensional space having 24 axes indicating the number of steps per hour is prepared. The group classification unit 303 classifies into groups using a space in which each axis corresponds to the number of steps taken by the user every hour from 0:00 to 24:00. Furthermore, the group classification unit 303 calculates the position of one point in space corresponding to the daily distribution pattern, and similarly calculates the positions of multiple points corresponding to the multiple distribution patterns that have been calculated. do. Therefore, the distribution pattern calculation unit 302 obtains as many points as the number of distribution patterns in space. The group classification unit 303 clusters a plurality of points into one or more groups based on the positions of the points distributed in space. As a result, the clustering unit can classify the distribution patterns into groups that are assumed to have similar distribution trends. Alternatively, a vector whose components are the number of steps taken by the user per hour may be considered, and this vector may be associated with points in a space spanned by a basis determined from the components. For example, this vector indicates a position vector, where the starting point of the position vector corresponds to the origin where the number of steps is zero for all hourly periods included in a day, and the ending point of the position vector corresponds to one distribution pattern. .

When the group classification unit 303 classifies groups by class classification, the group classification unit 303 classifies the groups based on, for example, whether the amount of activity in a specific time period is greater than or equal to a threshold value. For example, it is classified into two groups depending on whether the threshold value is exceeded during the time period from 11:00 PM to 2:00 AM the next day. Threshold values are set for each of a plurality of time periods, and the groups are divided depending on whether or not the threshold value is exceeded, thereby generating a plurality of groups. Alternatively, the group classification unit 303 may set a plurality of threshold values for the same time period, and group the objects into three or more based on the threshold values for this time period. For example, the group classification unit 303 classifies into three groups: a group that takes less than 500 steps between 10 a.m. and 12 a.m., a group that takes more than 500 steps but less than 1,000 steps, and a group that takes more than 1,000 steps. In this case, there are two threshold values: 500 steps and 1000 steps.

The external factor information acquisition unit 304 acquires external factor information regarding external factors that are not directly related to the user and are factors that affect the user, other than the user's attribute information, via the communication interface 201. The information is acquired from the network 140 or the like. External factor information includes, for example, location information indicating the user's current location, weather forecast information based on location information (e.g., weather conditions for each hour in the future), and schedule information for the user (e.g., weather conditions for each hour in the future). There is a schedule for each day. In this embodiment, this external factor information is information used to calculate a target pattern that is a goal for the user in association with a group. In addition to location information, it may also include location history information that records the history of this location, and may also record or predict the ease of doing activities depending on a specific location and include this in the external factor information. Good too. The external factor information may be any external factor brought to the user, and is not limited to the items described here. Furthermore, the external factor information may be used by the advice presentation unit 307 to present appropriate advice information.

The attribute information acquisition unit 305 acquires attribute information, which is information regarding user attributes, from the storage unit 202. In addition, attribute information is information used when calculating the user's target pattern, and includes, for example, physical information (e.g., gender, height, weight) necessary for measuring the user's activity level, external factors There is information on the place of residence to be associated with the user, occupational information such as occupation and place of work, information on hobbies related to activities, and information on eating and drinking preferences used when eating out. Further, the attribute information may be used by the advice presentation unit 307 to present appropriate advice information.

The target pattern calculation unit 306 calculates a target pattern that is a user's goal for each group classified by the group classification unit 303 based on the attribute information and external factor information. The target pattern calculation unit 306 calculates a target pattern desirable for the user according to the user's attribute information and external factors brought about to the user, while indirectly reflecting the characteristics of the group. Furthermore, the target pattern calculation unit 306 calculates future target patterns by also referring to the distribution patterns currently in progress and which have not yet been completed and/or the distribution patterns and target patterns (and/or advice information) from the previous day. It's okay.

The advice presenting unit 307 selects a target pattern corresponding to the currently ongoing distribution pattern from among the currently ongoing distribution pattern acquired from the distribution pattern calculating unit 302 and the target pattern calculated for each group by the target pattern calculating unit 306. Based on this, advice information for the user is generated and delivered to the output device 204. Advice information is generated every time a target pattern is calculated, and every time the target pattern is changed. Furthermore, the storage unit 202 may always store the latest advice information so that when the user performs a predetermined operation, the advice information based on the target pattern that is valid at that time can be presented to the user. When the output device 204 receives advice information from the advice presentation unit 307, it presents this information to the user. The pattern that is currently being acquired has content that corresponds to the target pattern, and the first and second periods for which the distribution pattern was acquired have the same length and the same type of activity as the target pattern. It is something. Examples of the types of activity amount include energy consumption, activity energy, and number of steps. The advice presentation unit 307 stores the distribution pattern and the target pattern for the second period in the storage unit 202 as history information. Further, the advice presentation unit 307 may also store advice information in the storage unit 202 in association with the distribution pattern and the target pattern. Further, the storage unit 202 may store the difference pattern between the distribution pattern and the target pattern in association with the advice information (for example, as a table).

<Others>
The operation of the habit improvement device 100 will be explained in detail in the operation example described later. In addition, in this embodiment, the control unit 205 of the habit improvement device 100 may be realized by a general-purpose CPU. However, some or all of the above operations (or functions) may be realized by one or more dedicated processors. Further, regarding the configuration of the habit improvement device 100, omissions, substitutions, and additions may be made as appropriate depending on the embodiment.

Next, using FIG. 4, an example of the user's daily behavior when using the habit improving device 100 and advice information of the habit improving device 100 will be displayed and explained. FIG. 4 is a time chart schematically showing an example of the user's daily behavior and an example of the operation and display of the habit improvement device 100.
(Midnight) The user plans to go to bed by midnight and wake up at 8 am. Even during sleep, the user wears the activity meter 110 and/or the wristwatch-type wearable terminal 120, these devices detect the activity amount, and the activity amount acquisition unit 301 of the habit improvement device 100 acquires this activity amount information. Further, when sleeping, the activity meter 110 and the wristwatch-type wearable terminal 120 may be turned off and the user does not wear them, and the habit improvement device 100 may also be turned off.

(8 a.m.; wake up) The user wakes up at 8 a.m., and then inputs the user's attribute information, external factor information, schedule information, etc. into the habit improvement device 100 using the input device 203. However, among these pieces of information, information that can be obtained in advance (for example, user attribute information, schedule information) may be stored by the user in the storage unit 202 of the habit improvement device 100 in advance. . Further, after the user wakes up, the output device 204 of the habit improvement device 100 is caused to display the user's target pattern for the day. The target pattern is calculated by the control unit 205 of the habit improvement device 100 based on the amount of activity up to waking up. In particular, if there is advice information associated with the target pattern, the advice information is displayed on the output device 204. In the example shown in FIG. 4, the target pattern is displayed along with the words "Today is this". Of course, the text may be output as audio by the output device 204. Further, advice information may be displayed or audibly presented to the user to tell the user to wear the activity meter 110 and/or the wristwatch-type wearable terminal 120 immediately after waking up or after a while. Thereafter, the user wears the activity meter 110 or the like, checks the target pattern, and starts the day's activities.

(12:00 AM; Interim Check) The habit improvement device 100 calculates a target pattern, and further generates advice information based on the distribution pattern indicating the amount of activity in the morning and the target pattern (Review (1)). The target pattern and the morning distribution pattern are displayed on the output device 204, and advice information is also presented to the user in the form of voice or image. In the example of FIG. 4, since the amount of activity in the morning was smaller than the amount of activity according to the target pattern, the output device 204 presents the user with advice information such as "You're not enough, so walk a little more later!" Further, the habit improvement device 100 presents the target pattern and the distribution pattern to the user using the output device 204.

(23:00 p.m.; final check) Similar to the intermediate check, the habit improvement device 100 calculates a target pattern, and further generates advice information based on today's distribution pattern and the target pattern (Review (2) ). The output device 204 displays today's target pattern and the distribution pattern indicating today's activity level, and also presents advice information to the user in the form of audio or images. For example, the habit improvement device 100 displays today's target pattern and today's distribution pattern as shown in FIG. Let's get to work early and walk!''), and when the target pattern calculation unit 306 calculates a target pattern from tomorrow onward, it calculates the target pattern based on these patterns and/or advice information as well.

(24:00 PM; Bedtime) The habit improvement device 100 may use the output device 204 to present tomorrow's target pattern to the user before the user goes to bed. It is expected that by knowing tomorrow's target pattern in advance, the user will be able to smoothly carry out tomorrow's activities. Further, before going to bed, the output device 204 may present to the user the advice information shown in the example of FIG. 4, "Let's go to work early tomorrow morning and walk."

[Operation example: Overall]
Next, an outline of an example of the operation of the habit improvement device 100 will be described using FIG. 5.

FIG. 5 is a flowchart illustrating an example of the processing procedure of the habit improvement device 100. Note that the processing procedure described below is only an example, and each process may be changed as much as possible. Further, regarding the processing procedure described below, steps can be omitted, replaced, or added as appropriate depending on the embodiment.

(boot)
First, the user starts up the habit improvement device 100 via the input device 203, and further receives inputs such as settings. The control unit 205 of the habit improvement device 100 advances the process according to the following processing procedure.

(Step S501)
In step S<b>501 , the control unit 205 operates as the attribute information acquisition unit 305 and acquires user attribute information from the storage unit 202 . This attribute information is stored in advance in the storage unit 202 via the input device 203 or the communication interface 201, for example. The user uses the input device 203 to store his or her own attribute information in the storage unit 202 . Alternatively, the user's attribute information may exist on the cloud, and the habit improvement device 100 may acquire the attribute information from there via the communication interface 201 and store it in the storage unit 202. Note that since the attribute information is used in step S504, step S501 may be executed at any time before step S504.

(Step S502)
In step S502, the control unit 205 operates as the activity amount acquisition unit 301, and acquires activity amount information (for example, step count data) from the activity meter 110 and/or wristwatch-type wearable terminal 120 via the communication interface 201. , the activity amount information is stored in the storage unit 202. The activity amount acquisition unit 301 obtains, for example, hourly activity amount information and stores this activity amount information in the storage unit 202 together with time zone information.

(Step S503)
In step S503, the control unit 205 operates as the distribution pattern calculation unit 302 and the group classification unit 303. The distribution pattern calculation unit 302 acquires the hourly activity amount information acquired by the activity amount acquisition unit 301 from the storage unit 202, and calculates a distribution pattern in which this activity amount is graphed. Then, the group classification unit 303 classifies the plurality of distribution patterns calculated by the distribution pattern calculation unit 302 into one or more groups. The calculated distribution pattern is then stored in the storage unit 202 in association with the group.
(Step S504)
In step S504, the control unit 205 operates as an external factor information acquisition unit 304, an attribute information acquisition unit 305, and a target pattern calculation unit 306. The external factor information acquisition unit 304 acquires the user's location information, external factor information, and user schedule information from the GPS receiving unit 208 and the communication interface 201. The attribute information acquisition unit 305 acquires user attribute information. The target pattern calculation unit 306 calculates a target pattern for each group based on the user's attribute information, location information, external factor information, and schedule information. The target pattern calculation unit 306 may further update the target pattern by comparing the calculated target pattern with the currently ongoing distribution pattern. That is, for example, based on the user's attribute information and external factor information, each time the number of steps is obtained, it is compared with the target pattern, and the target pattern is reviewed every hour to determine the target number of steps for the day. A target pattern for achieving a certain number may be generated.

(Step S505)
In step S505, the control unit 205 operates as the advice presentation unit 307 and compares the target pattern calculated by the target pattern calculation unit 306 with the latest distribution pattern calculated by the distribution pattern calculation unit 302 to achieve the goal. Advice information is generated according to the degree of severity, and the advice information is presented to the user via the output device 204.

[Operation example: S502]
Next, an example of step S502 in FIG. 5 will be described using FIGS. 6A and 6B.
(Step S601)
In step S601, the control unit 205 operates as the activity amount acquisition unit 301, and acquires activity amount information (for example, step count data here) from the activity meter 110 and/or the wristwatch-type wearable terminal 120 via the communication interface 201. do.

(Step S602)
In step S<b>602 , the control unit 205 operates as the activity amount acquisition unit 301 and causes the storage unit 202 to store information on the number of steps, which is the acquired activity amount. The activity amount acquisition unit 301 acquires activity amount information for each hour, for example, and stores this information in the storage unit 202. As shown in FIG. 6B, the activity amount acquisition unit 301 collects the user ID (User_ID) that can identify the user, the date, day of the week, and time when the amount of activity was detected by the activity meter 110 and/or the wristwatch type. Information acquired from the wearable terminal 120 and associated with the amount of activity (this is referred to as measurement information) is stored in the storage unit 202. The table shown in FIG. 6B includes an example of measurement information.

Note that when the activity meter 110 and/or the wristwatch-type wearable terminal 120 detect the activity amount and transmit it to the habit improvement device 100 almost simultaneously, the activity amount acquisition unit 301 transmits the date, day of the week, and time. The date, day of the week, and time at which the activity amount information was acquired via the interface 201 may also be used. In this case, the habit improvement device 100 uses the clock device 206 to obtain the year, month, and day.

[Operation example: S503]
Next, an example of step S503 in FIG. 5 will be described using FIG. 7A.
(Step S701)
In step S701, the control unit 205 operates as the distribution pattern calculation unit 302, and the distribution pattern calculation unit 302 acquires measurement information including the hourly activity amount information acquired by the activity amount acquisition unit 301 from the storage unit 202. Then, a distribution pattern that can be used to graph this measurement information is calculated.

(Step S702)
In step S702, the control unit 205 operates as the group classification unit 303, and the group classification unit 303 classifies the plurality of distribution patterns calculated by the distribution pattern calculation unit 302 into one or more groups. Then, the group classification unit 303 stores the calculated distribution pattern (for example, hourly step count data for one day) in the storage unit 202 in association with the group. The group classification unit 303 classifies a plurality of distribution patterns in a specific period in the past. The distribution pattern is daily step count data, and the group classification unit 303 classifies the step count data for the past month, for example.

[Operation example: S701 and S702]
Next, an example of detailed operations in steps S701 and S702 in FIG. 7 will be described using FIG. 7B.
(Step S711)
In step S711, the control unit 205 operates as the distribution pattern calculation unit 302, and transmits hourly user activity information from the activity meter 110 and/or wristwatch wearable terminal 120 to the storage unit 202 via the communication interface 201. Get from.

(Step S712)
In step S712, the control unit 205 operates as the distribution pattern calculation unit 302, and converts the activity amount information acquired in step S711 into the date and time when the activity meter 110 and/or the wristwatch-type wearable terminal 120 detected the activity amount, or the habit A distribution pattern is calculated according to the date and time when the improvement device 100 acquired the amount of activity. As a result, for example, from one day's worth of activity, a distribution pattern indicating a change over time in the amount of activity with respect to the time of the day is calculated.

(Step S713)
In step S713, the control unit 205 operates as the group classification unit 303 and classifies the plurality of distribution patterns calculated by the distribution pattern calculation unit 302 into one or more groups. For example, the group classification unit 303 maps the distribution pattern of the amount of activity for 24 hours onto a 24-dimensional space. In this 24-dimensional space, each axis corresponds to the amount of activity per hour. In other words, this space is a 24-dimensional space defined by a basis determined from vectors whose components are hourly amounts of activity (24 pieces). Therefore, the distribution pattern of the amount of activity on a given day corresponds to one point in the 24-dimensional space. That is, a plurality of points corresponding to all the distribution patterns calculated by the distribution pattern calculation unit 302 are distributed in the 24-dimensional space. Even if only one of the activity amounts corresponding to each axis differs, the positions of the points in the 24-dimensional space are different, and the closer the distance in the 24-dimensional space, the more the distribution pattern between these points will change. are considered to be similar.

(Step S714)
In step S714, the control unit 205 operates as the group classification unit 303 and determines how to group the points mapped in the 24-dimensional space. In the example of FIG. 7B, it is determined whether the points mapped by the group classification unit 303 are to be grouped using a clustering method. This determination is usually made by determining in advance whether the grouping is to be performed using a clustering method or not when the user or the habit improvement device 100 is designed. The user may be able to set whether to use clustering using the input device 203 of the habit improvement device 100, and in step S714 it may be determined whether to perform clustering based on this setting.

(Step S715)
In step S715, the control unit 205 operates as the group classification unit 303, groups points distributed in the 24-dimensional space using a clustering method, and generates one or more groups. Here, the explanation assumes hierarchical clustering, which is a clustering method, but it is not limited to this method, and split optimization clustering, etc. may also be used (however, there will be changes in the procedure) .

(Step S716)
In step S716, the control unit 205 operates as the group classification unit 303 and determines whether the number of groups generated in step S715 corresponds to a desired hierarchy of clusters. If it is determined in step S716 that it corresponds to the desired hierarchy, the process advances to step S717; on the other hand, if it is determined in step S716 that it does not correspond to the desired hierarchy, the process returns to step S714. In the case of agglomerative hierarchical clustering, each point in space is an individual cluster, then these clusters are successively merged to generate a hierarchy of clusters, and finally merged into a single cluster. do. Here, the hierarchy refers to a state in which two clusters are merged, and the hierarchy changes as the merging progresses. For example, if the cluster hierarchy is different, the number of points included in the cluster will be different.

In the case of split-type hierarchical clustering, which is the opposite of agglomerative hierarchical clustering, it is clustering that divides from one cluster, and the division starts from a state where the entire data collection is one cluster. Then, these are divided sequentially to generate a hierarchy of clusters.

Further, in step S716, the group classification unit 303 may determine whether the number of groups has reached a desired number. If it is determined that the number of groups has reached the desired number, the process advances to step S717, whereas if it is determined that the number of groups has not yet reached the desired number, the process advances to step S714.

(Step S717)
In step S717, the control unit 205 operates as the group classification unit 303, and determines the number of groups corresponding to the desired hierarchy determined in step S716 and the groups corresponding to this number of groups.

(Step S718)
In step S718, the control unit 205 operates as the group classification unit 303, and each point in the space is classified into one of the groups determined in step S717.

(Step S719)
In step S719, the control unit 205 operates as the group classification unit 303 and groups one or more points distributed in the 24-dimensional space using a method different from clustering (for example, class classification). As an example, these time periods are extracted in order to classify based on the time period when the amount of activity is high (and/or low). The method from step S719 onwards differs from the clustering method in that the characteristics of the group are set in advance and points in the space are classified based on the characteristics.

(Step S720)
In step S720, the control unit 205 operates as the group classification unit 303 and generates one or more groups having characteristic time periods of activity amount. For example, if the user has set that weekdays are considered to have the same distribution pattern, the group classification unit 303 uses the activity amount during the daytime as the first threshold as the group classification criterion. This is the above, and if the amount of activity in other time periods is less than or equal to the second threshold, it is assumed that the objects belong to the same group. In addition, different from this, for example, the average value (for example, variance and standard deviation may also be calculated) for each time period on weekdays is calculated from data on past distribution patterns of users, and the group classification unit 303 indicates that if a point corresponding to a certain distribution pattern is located in a space centered on the average value and having a confidence level of 95%, this distribution pattern belongs to this weekday group. It may be classified as

The groups generated by the group classification unit 303 correspond to specific regions in space (for example, 24-dimensional space), and the groups do not intersect with each other. Furthermore, the group classification unit 303 can create groups having various other characteristics. For example, the group classification unit 303 can generate a corresponding group by setting a classification standard that captures the characteristics of the amount of activity on holidays in the same manner as described above.

(Step S721)
In step S721, the control unit 205 operates as the group classification unit 303 and classifies points in space into one or more groups generated in step S720.

(Step S722)
In step S722, the control unit 205 operates as the group classification unit 303, and determines that each of all points distributed in the space belongs to one of the groups generated in step S721, and that unclassified points are Determine whether or not there is. If the group classification unit 303 determines that there is a point that is not classified into a group, the process advances to step S723, whereas if it determines that there is no point that is not classified to a group, the process advances to step S718.

(Step S723)
In step S723, the control unit 205 operates as the group classification unit 303, and selects any of the groups generated in step S720 within a preset distance from the point determined to be unclassified into any group. If two or more groups match, the unclassified point is classified into the closest (smallest distance) group, and if one group matches, it is classified into that group. Classify unclassified points.

There are various definitions for the distance from a point in space to a group, and any definition may be used as long as it is uniquely determined and reasonable. For example, this distance may be the minimum distance from a point in space to a point belonging to the desired group. Alternatively, a group boundary may be defined, and the distance from a point in space to the group boundary may be defined as the distance between the point and the group. Here, the former definition is adopted, and in step S723, the group classification unit 303 calculates the distance between the point that is determined to be unclassified into any group and the point that belongs to a certain group. A point with a small distance is searched, and it is determined whether this distance is within a preset distance. If this distance is within a preset distance, the point determined to be unclassified belongs to this group, and it is determined whether there are any other unclassified points.

(Step S724)
In step S724, the control unit 205 operates as the group classification unit 303, determines whether there are any unclassified points larger than the preset distance in step S723, and determines whether there are any unclassified points larger than this distance. If it is determined that there is no unclassified point, the process advances to step S718, whereas if there is an unclassified point with a distance greater than this distance, the process advances to step S725. Since unclassified points do not belong to any group, it is necessary to generate a new group that includes these unclassified points.

(Step S725)
In step S725, the control unit 205 operates as the group classification unit 303, generates a group to include the points determined as unclassified points in step S724, and returns to step S721. This generated group is, for example, a multidimensional sphere (e.g., , 24-dimensional sphere), and the points existing in the area within this multidimensional sphere may be in the same group. If groups are generated in this way, the generated groups will not overlap in area with other groups.

Note that here, the method for grouping one or more points in space is an example in which only one of a clustering method and a method not based on clustering such as class classification is used. , both techniques may be used. For example, multiple groups that define the amount of activity for one or more specific time periods are set in advance, and points existing in spatial areas that do not belong to these groups are classified into one or more groups by clustering. There is a method.

Next, using FIG. 7C, a specific example of the result of classifying distribution patterns into groups using, for example, the flowchart shown in FIG. 7B is shown. FIG. 7C shows the distribution pattern of the number of steps for three weeks on a daily basis, and this example shows a case where the distribution pattern is classified into four groups.

In FIG. 7C, different groups are shown with different bar graph patterns. In the example of FIG. 7C, (1) a bar graph with an outline surrounded by line segments and a hatched interior; (2) a bar graph with an outline surrounded by broken lines and a white interior; ) bar graphs with outlines surrounded by line segments and black inside; (4) bar graphs with outlines surrounded by line segments and white inside. ing. Note that this example does not guarantee that the classification result will be exactly the same whether it is classified using clustering, class classification, or any other method, but almost the same result can be obtained using any of the methods.

The bar graphs for five weekdays surrounded by a solid line frame 751 have the same weekday category, but the distribution patterns are different, so the distribution pattern from Monday to Thursday (corresponding to (1)) and Friday's are different. The distribution pattern (corresponding to (2)) is classified as belonging to a different group. This grouping is determined, for example, based on whether the amount of activity is greater than a predetermined value in a preset time period as shown in steps S719, S720, and S721. In this case, one of the criteria for determination is, for example, whether there is an amount of activity (number of steps) after 11:00 p.m. That is, on Fridays within the solid line frame 751, the number of steps after 11:00 PM is 500 or more, and on Mondays to Thursdays within the solid line frame 751, the number of steps after 11:00 PM is less than 500 steps (in this example, 0 steps). It is possible that this classification can be similarly performed by clustering if the cluster hierarchy that determines the number of groups is adjusted.

The bar graphs for four days surrounded by a dotted line frame 752 have the same days of the week, but have different distribution patterns, so it can be seen that even though the days of the week are the same, they are classified as belonging to different groups. This grouping can also be performed in the same manner as the grouping within the solid line frame 751 described above. Note that, for example, the upper left graph and the lower right graph in the dotted line frame 752 have a distribution pattern in which the number of steps after 11:00 pm is less than 500 (corresponding to (1)), and the upper right graph in the dotted line frame 752 The lower left graph shows a distribution pattern (corresponding to (2)) in which the number of steps taken after 11 p.m. is all 500 or more.

The bar graphs for the 6 days surrounded by the double-dashed line frame 753 are the same in terms of Saturdays and Sundays, but since the amount of activity in specific time periods differs between Saturday and Sunday, the distribution pattern for Saturday ((3) (corresponding to (4)) and the Sunday distribution pattern (corresponding to (4)) are classified as belonging to different groups. This grouping can also be performed in the same manner as the grouping within the solid line frame 751 described above. For example, on Saturday, the number of steps is 100 or more after 0:00, and the total number of steps from 10:00 to 13:00 is 1,000 or more, while on Sunday, the number of steps is less than 100 (0 steps) after 0:00. Yes, and the total number of steps from 10:00 to 13:00 is less than 1,000 steps.

[Operation example: S504]
Next, an example of step S504 in FIG. 5 will be described using FIG. 8A.
(Step S801)
In step S801, the control unit 205 operates as the external factor information acquisition unit 304 and the attribute information acquisition unit 305, and stores the external factor information of the user and the user attribute information for each user in the storage unit 202 and the GPS reception unit. 208 and/or the communication interface 201 .

(Step S802)
In step S802, the control unit 205 operates as the target pattern calculation unit 306, and calculates a typical pattern for each group from the distribution patterns classified into groups in step S503 according to a classification method such as clustering or class classification. A pattern (or basic pattern) is identified, and a target pattern for each user is calculated for each group from this typical pattern based on the attribute information and external factor information for each user acquired in step S801. Note that a pattern that is typical for a group when grouped by methods such as clustering or class classification is called a typical pattern; however, in the embodiment, for convenient identification, the number of distribution patterns to be samples is set to a threshold value. A typical pattern for a group when grouped in the following cases will be referred to as a basic pattern.

[Operation example: S802]
Next, an example of step S802 in FIG. 8A will be described using FIG. 8B.
(Step S811)
In step S811, the control unit 205 operates as the target pattern calculation unit 306, and determines whether the number of distribution patterns calculated in step S503 is larger than a threshold value. This threshold is a criterion for determining whether the number of distribution patterns is large enough to classify the characteristics of a group, and is usually a preset value. It is assumed that the larger the threshold value is, the larger the number of distribution patterns for which a YES determination is made in step S811, and the more prominent the characteristics of each group will be, making it possible to obtain a more accurate target pattern. For example, when acquiring the distribution pattern of the number of steps every day as shown in FIG. There is an example in which the process proceeds to S812.

In step S812, the control unit 205 operates as the target pattern calculation unit 306, and identifies a typical pattern that is typical of the group for each group obtained by clustering, class classification, and the like. In other words, in this embodiment, the larger the number of distribution patterns, the more accurately groups can be classified by clustering etc., and the more accurate the target pattern can be obtained. If this is also large, a typical pattern is obtained for each group obtained by a method such as clustering. To specify the typical pattern, the target pattern calculation unit 306 calculates, for example, a distribution pattern corresponding to a point at the center or center of gravity of the group from the patterns belonging to each group as the typical pattern. Alternatively, the target pattern calculation unit 306 may randomly select a pattern from the distribution patterns belonging to a group and use it as the typical pattern of the corresponding group.

In step S813, the control unit 205 operates as the attribute information acquisition unit 305 and the target pattern calculation unit 306, and acquires the user's attribute information from the storage unit 202.

In step S814, the control unit 205 operates as the external factor information acquisition unit 304 and the target pattern calculation unit 306, and acquires external factor information from the communication interface 201 and the GPS reception unit 208.

In step S815, the control unit 205 acquires the user's attribute information in the same manner as in step S813.

In step S816, the control unit 205 identifies one or more groups corresponding to the attribute information acquired in step S815. For example, the information includes the user's physical information, place of residence, hobby information, and occupational information such as occupation and place of work. In addition to this, for example, if the user is Japanese and an office worker, weekdays, Saturdays, and Sundays are assumed to be in different groups, so the user's attribute information and weekdays, Saturdays, and Sundays are divided into three groups. But that's fine. Furthermore, if the user has lessons or regularly goes to a gym, etc., the number of groups may be further increased by taking these into consideration.

In step S817, the control unit 205 operates as the target pattern calculation unit 306, selects one pattern from the distribution patterns assumed for each group identified in step S816, sets this as a basic pattern, and selects one pattern for each group. Identify basic patterns. The assumed distribution pattern may be freely set by the user as long as it belongs to a group, and is set according to the user's attribute information. Further, the target pattern calculation unit 306 may automatically generate distribution patterns according to the user's attribute information, and each of these distribution patterns may be used as the basic pattern of the group.

In step S818, the control unit 205 operates as the target pattern calculation unit 306, and generates the user's target pattern from the user's typical pattern or basic pattern based on the user's attribute information and external factor information.

The specific steps of FIG. 8B will now be described with reference to FIGS. 8C and 8D. In FIG. 8C, typical patterns for each group obtained by classifying distribution patterns or basic patterns for each group set based on user attribute information, etc., and external factor information from the typical patterns or basic patterns are shown. The target pattern generated for each user is shown. FIG. 8D shows an example in which the amount of activity is increased or decreased or the activity time period is shifted from the typical pattern or the basic pattern based on the user's schedule information.

The typical pattern identified in step S812 or the basic pattern identified in step S817 is, for example, the distribution pattern shown on the left side of FIG. 8C. That is, the target pattern calculation unit 306 specifies a typical pattern or basic pattern for a certain day based on the user's attribute information.

Then, the target pattern calculation unit 306 acquires external factor information as in steps S814 and S818, and generates a target pattern based on this information.
In the example of FIG. 8C (a), the target pattern calculation unit 306 increases or decreases the amount of activity based on the user's activity range, such as the user's place of residence and place of work. In (a), it is known that there is a lot of precipitation during a specific time period in the area of this user's activity range, so the amount of activity (in this example Then, reduce the number of steps) from the basic pattern or typical pattern. Then, this reduced amount of activity is allocated to another time period (in this example, from 3:00 pm to 6:00 pm).

In the example shown in FIG. 8C (b), based on the schedule information of the external factor information, since this day is an early work day, the basic pattern or the typical pattern as a whole is brought forward to match the work start time. Then, the target pattern calculation unit 306 generates a target pattern for this day based on the basic pattern or typical pattern that has been brought forward, and further based on other external factor information and attribute information.

In the example of FIG. 8D, the target pattern calculation unit 306 changes the activity time period and activity amount according to the user's schedule information based on today's basic pattern or typical pattern shown on the left side of the figure. The example in the upper right of FIG. 8D is a case where a banquet is scheduled from 7:00 pm today, and the target pattern calculation unit 306 calculates that the activity amount at the time of the banquet will decrease, so Generate a target pattern to increase the number of steps by walking extra or jogging lightly during commuting.

In the lower right example of FIG. 8D, the person is scheduled to exercise at the gym from 7:00 pm today, so the target pattern calculation unit 306 generates a target pattern that increases the amount of activity at the gym. Further, since there is no need to reduce the amount of activity other than at the gym, the target pattern calculation unit 306 generates a target pattern by assuming that other activities follow the basic pattern or typical pattern and there is no problem.

Next, using FIG. 8E, it will be explained that when the user takes an action different from the scheduled information, the target pattern calculation unit 306 changes the target pattern according to this action.
The target pattern calculation unit 306 determines whether the user is following the schedule properly, and if the user is performing an action different from the schedule, changes the target pattern based on the action. For example, the target pattern calculation unit 306 acquires the current amount of activity from the activity amount acquisition unit 301, acquires the user's location information and schedule information from the external factor information acquisition unit 304, and monitors the user's behavior.

Further, as for the currently ongoing, recent, and/or current activity amount information, the activity amount information acquired by the activity amount acquisition unit 301 is stored in the storage unit 202, and the activity amount information is transferred to the storage unit 202 to achieve the goal. The pattern calculation unit 306 may obtain the information. The target pattern calculation unit 306 monitors the degree to which the user has achieved the target pattern, and if there is a deviation between the target pattern and the actual distribution pattern, changes the target pattern to ensure the amount of activity. You may. Further, even if the user suddenly changes his/her schedule, the target pattern calculation unit 306 may constantly monitor the schedule and amount of activity. Note that a case where the target pattern and the actual distribution pattern deviate from each other is, for example, a case where the difference between the target pattern, the actual distribution pattern, and the amount of activity in a certain time period is greater than a threshold value.

In the example at the top of FIG. 8E, the user was planning to go to the gym from 7:00 pm, but his mood changed and he ended up playing pachinko. In this case, the target pattern calculation unit 306 determines that the user is not acting according to the schedule based on the user's location information and schedule information provided by the GPS reception unit 208. In this way, it is also possible for the target pattern calculation unit 306 to automatically detect that the user is playing pachinko based on the position information, the user's activity amount information, and the like. In this case, the target pattern calculation unit 306 performs rescheduling such as changing the schedule from the gym to pachinko and incorporating the decreased amount of activity into tomorrow's schedule, for example.

In the example at the bottom of FIG. 8E, the user is scheduled to have a party starting at 7:00 pm, and is going to the party as planned. Based on the user's location information and schedule information from the GPS receiving unit 208, the target pattern calculating unit 306 confirms that the user is acting according to the schedule. In this case, since the user is acting according to the schedule, the target pattern calculation unit 306 determines that there is no need to change the target pattern.

[Operation example: S505]
Next, an example of step S505 in FIG. 5 will be described using FIG. 9A.
(Step S901)
In step S901, the control unit 205 operates as the advice presentation unit 307, and receives the currently ongoing distribution pattern from the distribution pattern calculation unit 302 (for example, the distribution pattern from midnight today to the present) and the information generated in step S802. A difference pattern is calculated by taking the difference from the target pattern. This difference pattern is a pattern in which the difference between the amount of activity in the distribution pattern for each time and the amount of activity in the target pattern is distributed over time.

(Step S902)
In step S902, the control unit 205 operates as the advice presentation unit 307, extracts advice information from the storage unit 202 from the table in which advice information corresponding to the difference pattern calculated in step S901 is written, and outputs the advice information from the output device 204. to the user via the

[Operation example: S901 and S902]
Next, an example of step S901 and step S902 in FIG. 9A will be described using FIG. 9B.
(Step S911)
In step S911, the control unit 205 operates as the advice presentation unit 307, and calculates the currently ongoing distribution pattern calculated by the distribution pattern calculation unit 302 from the activity amount information acquired by the activity amount acquisition unit 301 via the communication interface 201. get. This ongoing distribution pattern is, for example, today's distribution pattern, and is the distribution pattern from the time when the activity amount was first obtained today until now.

(Step S912)
In step S912, the control unit 205 operates as the advice presentation unit 307 and obtains today's target pattern from the target pattern calculation unit 306. The target pattern calculation unit 306 calculates a target pattern based on the user's today's external factor information, attribute information, and the like.

(Step S913)
In step S913, the control unit 205 operates as the advice presentation unit 307, and calculates the difference between the distribution pattern currently in progress today and the target pattern calculated in step S912 within a time period (first period; for example, 1 hour). The difference in activity amount for each time period is calculated, and a difference pattern is calculated by arranging these differences across all time periods.

(Step S914)
In step S914, the control unit 205 operates as the advice presentation unit 307, and generates advice information for the user based on the difference pattern and/or external factor information calculated in step S913. This advice information is stored in the storage unit 202, for example, in association with the difference pattern. Further, the advice presentation unit 307 may generate advice information based on the state of the difference pattern. Advice information is provided, for example, when the difference pattern shows a negative value (more specifically, when the cumulative amount of activity in the target pattern is smaller than the cumulative amount of activity in the corresponding time period of the distribution pattern). is a message such as "I'm running a little short, so walk a little more later!" and is presented to the user by display or voice. As for other advice information, for example, if a banquet is scheduled in the schedule information from 19:00 today, an example of advice that is presented to the user in the morning after waking up is ``Cut down the amount you drink and take more walks in the morning.''", and as another example, if the schedule information indicates that the user is scheduled to go to the gym from 7:00 pm today, there is an encouraging message that recommends exercising at the gym.

Further, the advice presentation unit 307 may generate advice information for the user based on external factor information. Among the external factor information, it is mainly assumed that advice information is presented using, for example, the user's schedule information, the user's location information, and weather forecasts based on the schedule information. The target pattern calculation unit 306 passes this external factor information to the advice presentation unit 307 along with the reason why the information referred to when calculating the target pattern by the target pattern calculation unit 306 is referred to. For example, if the weather forecast for today's scheduled run time is rainy, the target pattern calculation unit 306 lowers the planned activity amount during the scheduled run time and increases the activity amount during other times. Calculate the target pattern as follows. In this case, the target pattern calculation unit 306 passes this information to the advice presentation unit 307, and the advice presentation unit 307 generates advice information corresponding to this information and presents it to the user.

Another example of advice information based on schedule information is when a party is scheduled from 7:00 pm today, or when you are planning to go to the gym from 7:00 pm today, etc. When the target pattern is calculated based on this schedule information, reflection information indicating how the schedule information is reflected in the target pattern is passed to the advice presentation unit 307. The advice presentation unit 307 generates advice information based on this reflected information and presents it to the user.

Next, using FIG. 9C, an example of the advice information presented by the advice presentation unit 307 will be shown by showing the target pattern calculated by the target pattern calculation unit 306 and the actual distribution pattern currently in progress for today.

The target pattern calculation unit 306 acquires the distribution pattern of the amount of activity until 18:00 from the storage unit 202 or the activity amount acquisition unit 301, and further based on the information from the group classification unit 303 and the external factor information acquisition unit 304, A target pattern at the present moment (in the example of FIG. 9C, 18:00) is created. The target pattern in Figure 9C recommends that you exercise by returning home early and going to the gym from 8:00 pm, since the amount of activity up to this point is less than the amount of activity in the target pattern (reflected information). . The goal pattern calculation unit 306 passes this reflected information to the advice presentation unit 307, and the advice presentation unit 307, based on this reflected information, makes a message such as, “Start at 8 pm, exercise a little harder! Let's leave work early today.” generate advice information and present it to the user. Further, the reflection information and advice information may be stored in the storage unit 202 in association with each other, and the advice presentation unit 307 may acquire the advice information from the storage unit 202 based on the reflection information received from the target pattern calculation unit 306. .

[Action and effect]
As described above, the habit improvement device 100 of this embodiment records the hourly step count information acquired in step S502, and calculates the distribution pattern of the daily step count based on the step count information in step S503. Then, the habit improvement device 100 classifies the plurality of distribution patterns calculated by the distribution pattern calculation unit 302 into groups by clustering and/or class classification in the group classification unit 303, and further in step S504, calculates a target pattern. In section 306, a target pattern that is a desirable target for each group of users is calculated based on user attribute information and external factor information. This target pattern is calculated based on the user's attribute information and external factor information. Therefore, after classifying the user's characteristics by group, the user's attribute information, which is not easy to extract from the distribution pattern alone, is included, so the habit improvement device 100 is able to calculate a target pattern that is effective for the user. can. Furthermore, the target pattern is calculated based on the user's external factor information, and is calculated based on external factors such as the environment other than the user himself/herself. For this reason, calculations include external factor information that cannot be extracted from the distribution pattern alone, or even by combining the user's attribute information and the distribution pattern, thereby improving habits. The device 100 can further calculate a highly accurate and effective target pattern for the user.

Further, in step S505, the advice presentation unit 307 can present appropriate advice information to the user based on the calculated target pattern, for example, in association with the user's schedule information. Further, the target pattern calculation unit 306 can modify and/or change the target pattern at any time based on the currently ongoing distribution pattern and the target pattern corresponding to this distribution pattern, and modify the target pattern. And/or in the case of a change, the advice presentation unit 307 presents advice information corresponding to the modification change to the user. As a result, according to the habit improvement device 100, the user can receive advice information regarding activities at an appropriate timing, and can increase the possibility of achieving the target pattern.

[Modified example]
Although the embodiments of the present invention have been described in detail above, the above descriptions are merely illustrative of the present invention in all respects. It goes without saying that various improvements and modifications can be made without departing from the scope of the invention. For example, the following changes are possible. Further, in implementing the present invention, specific configurations depending on the embodiments may be adopted as appropriate. In addition, below, the same code|symbol is used regarding the same component as the said embodiment, and description is abbreviate|omitted suitably about the same point as the said embodiment. The following modifications can be combined as appropriate.

<1>
The external factor information acquisition unit 304 acquires attribute information of a user other than the user (as described above, for example, information such as age, place of residence, and/or gender) and the target pattern of the user, The attribute information of the user himself/herself acquired by the attribute information acquisition unit 305 is matched with the attribute information of the other user (for example, executed by the target pattern calculation unit 306), and if the degree of matching is higher than a preset criterion, may adopt this other user's target pattern.

Therefore, by referring to the attribute information of similar other users, it is possible to obtain the same effect as when sampling a large number of similar distribution patterns, so that it is possible to accurately calculate the user's target pattern. As a result, the habit improvement device of this embodiment can present more appropriate advice information to the user.

Further, in this case, if there is an actual sample in which another user has achieved the target pattern, it is desirable to make it easier to adopt that target pattern (for example, increase weighting). It can be expected that this goal pattern will be more likely to be achieved by the user than by adopting other unachieved goal patterns.

<2>
Referring to the distribution pattern and schedule information of other users with similar attribute information, the user uses the distribution pattern and/or target pattern of the other user in order to approach the attributes that the other user has already achieved. It is conceivable that desired attributes can be efficiently achieved by using To do this, first determine the attributes that the user wants to achieve, and then use other users who already have the attributes they want to achieve (for example, blood pressure value, BMI value), and whose other attributes are similar to the attributes of the user. Find out from a database etc. In order to find out the attributes of this other user, for example, the attribute information acquisition unit 305 connects to the server 130 via the communication interface 201 and performs a search. In this case, the server 130 may store attribute information, external factor information, and information on distribution patterns and target patterns from a large number of users.

Although it is effective for the user to simply view and refer to the distribution pattern and/or target pattern searched and found in this way, it is also effective to use this pattern as the target pattern as shown in the above embodiment. The target pattern calculation unit 306 may calculate the target pattern, and then the advice presentation unit 307 may present the advice information to the user.

If such a system were available, it would be possible to determine the actual distribution pattern of what kind of distribution pattern should be achieved in order for the user to achieve the desired attributes (for example, the user would achieve the ideal body shape). (distribution pattern of other users) can be used as a guideline, and it is more likely that the target pattern to be realized can be calculated with high accuracy.

In addition, the difference between the user's attribute information before the user realizes the target pattern and the user's attribute information after the user realizes the target pattern is stored in the server etc. for each user as attribute information for each item of attribute information. It is desirable to be present. In this case, by referring to the server and searching for similar attribute information of other users based on the current attribute information of the principal user and the desired attribute information, the principal user can be found from the same attribute information as the principal user. It becomes possible to efficiently find other users who have desired attribute information. This allows the user to efficiently find and use effective target patterns of other users, making it easier for the user to reach desired attributes. According to this configuration, for example, people in their 40s who want to be thin have many such lifestyle habits (corresponding to attribute information and distribution patterns and/or goal patterns), so they have different exercise habits (corresponding to distribution patterns). ) can be appropriately presented to the user.

<3>
In steps S719 and subsequent steps in FIG. 7B, points in the space may be classified into groups as follows, for example.

For example, if it is found from the user's attribute information that the user spends a specific time at a specific place for work on weekdays, the group classification unit 303 estimates that weekdays show almost the same distribution pattern, and The average value for each is calculated, and the distribution pattern having the average value is set as one of the basic patterns. The group classification unit 303 can also create basic patterns having various other characteristics. The group classification unit 303 can generate basic patterns for holidays in the same manner as described above.

For example, the group classification unit 303 may create groups and classify distribution patterns as described below, without introducing basic patterns. For example, the group classification unit 303 determines that the amount of activity is at a first threshold value (for example, a value that is 50% higher than the average value of the amount of activity on weekdays) late at night on weekdays (for example, from after 11:00 p.m. to before 2:00 the next day). Divide into groups based on whether there are more than Many criteria are set for the amount of activity and its time period, and these criteria are used simultaneously to classify into multiple groups. For example, there are two possibilities: whether the total amount of activity is smaller than the second threshold (for example, 50% of the amount of activity on weekdays), and whether the amount of activity late at night on weekdays is larger than the first threshold. When classifying based on criteria, it is possible to classify into 2×2=4 groups based on the first and second thresholds. In addition to these activity level criteria, other criteria may be introduced and points in space may be classified using these criteria as well. For example, a criterion of whether it is a weekday (Monday to Friday) or a Saturday, Sunday, or holiday may be introduced. Furthermore, if it is known that behavior on Fridays is different from weekdays, Fridays may be classified into a different group from weekdays.

<4>
The group classification unit 303 may classify into groups based on a plurality of preset typical patterns, which are typical patterns of groups. In this configuration, the group classification unit 303 classifies the plurality of distribution patterns into groups using a preset typical pattern indicating the characteristics of each group. The feature is that the typical pattern is set in advance, and the user or other setter can set the typical pattern in advance. Therefore, when classifying a plurality of distribution patterns of users into groups, it is possible to realize grouping with clear meanings.

<5>
Although not illustrated, the activity meter 110, the wristwatch-type wearable terminal 120, and/or the habit improvement device 100 may include an acceleration sensor, a pressure sensor, a gyro sensor, and/or a geomagnetic sensor.

Note that the acceleration sensor is a sensor that detects acceleration, and is, for example, a three-axis acceleration sensor, and detects the acceleration of the sensor with respect to three linearly independent axes (for example, three axes orthogonal to each other). The acceleration sensor outputs acceleration signals representing acceleration in three directions to the control unit 205. The acceleration sensor can obtain the roll angle and pitch angle from the value of acceleration when the vehicle is at rest.

A pressure sensor is a general device that detects pressure. For example, a pressure sensor can detect a user's altitude by measuring atmospheric pressure.

A gyro sensor is a general type that can detect the angular velocity of the sensor, and is, for example, a 3-axis gyro sensor, which detects the angular velocity of the sensor with respect to three linearly independent axes. The gyro sensor outputs angular velocity signals representing angular velocities in three directions to the control unit 205.

A geomagnetic sensor is common and is used to determine the user's posture. The geomagnetic sensor is, for example, a three-axis geomagnetic sensor, and detects the intensity of geomagnetism around the sensor along three axes, including direction and strength (size).

The control unit 205 obtains the angle by integrating the angular velocity from the initial posture information of the user, using a method of correcting the error due to drift in the angular velocity obtained from the gyro sensor, from the information of the acceleration sensor and the geomagnetic sensor. , the attitude angle for each sensor at a desired elapsed time from the initial time may be obtained. As for the initial posture, the roll angle and pitch angle can be obtained by an acceleration sensor. The yaw angle can be determined using a geomagnetic sensor.

A three-dimensional component of the magnetic field with the tilt error corrected is calculated from the three-dimensional component of the magnetic field obtained by the geomagnetic sensor and the previously determined roll angle and pitch angle. The yaw angle can be calculated from the x and y components of the three-dimensional components of the magnetic field that have been corrected for this tilt error. When the user moves from the initial posture, the angle is obtained by time-integrating the angular velocity, and the posture angle for each sensor at any time can be obtained from the initial posture and elapsed time.

Based on changes in the posture angle, movements of the user (e.g. movements of arms and/or legs) can be detected when the user is stationary, e.g. For example, dancing, exercising at a gym) can be detected more accurately than when these are not detected.

<6>
The apparatus of the present invention can also be realized by a computer and a program, and the program can be recorded on a recording medium (or a storage medium) or can be provided through a network.
Furthermore, each of the above devices and their device parts can be implemented either in a hardware configuration or in a combination configuration of hardware resources and software. The combined software is installed in advance on a computer from a network or a computer-readable recording medium (or storage medium), and is executed by the processor of the computer, thereby controlling the operations (or functions) of each device on the computer. A program is used to make this happen.

<7>
It should be noted that the present invention is not limited to the above-described embodiments as they are, but can be embodied by modifying the constituent elements within the scope of the invention at the implementation stage. Moreover, various inventions can be formed by appropriately combining the plurality of components disclosed in the above embodiments. For example, some components may be deleted from all the components shown in the embodiments. Furthermore, components from different embodiments may be combined as appropriate.

Further, "and/or" means any one or more of the items listed with "and/or". To give a specific example, "x and/or y" means any element of the set {(x), (y), (x, y)} consisting of three elements. To give another specific example, "x, y, and/or z" means a set of seven elements {(x), (y), (z), (x, y), (x, z ), (y, z), (x, y, z)}.

(Additional note 1)
A first calculation unit ( 301, 302) and
a classification unit (303) that classifies the plurality of distribution patterns into one or more groups;
a second calculation unit (304, 305, 306) that calculates a target pattern that is a goal for the user in association with the group, based on factors other than the amount of activity;
a presentation unit (307) that presents advice information to the user based on the current distribution pattern and the target pattern;
Habit improvement device (100).

100...Habit improvement device 110...Activity meter 120...Wristwatch type wearable terminal 130...Server 140...Network 150...GPS satellite 201...Communication interface 202...Storage unit 203...Input device 204...Output device 205...Control unit 206...Timekeeping device 207...Power source unit 208...GPS receiving unit 209...External interface 301...Activity amount acquisition unit 302...Distribution pattern calculation unit 303...Group classification unit 304...External factor information acquisition unit 305...Attribute information acquisition unit 306...Target pattern calculation unit 307...Advice presentation section 751...Solid line frame 752...One-dot broken line frame 753...Two-dot broken line frame

Claims (12)

A first calculation unit that acquires the amount of activity of the user in each of a plurality of first periods constituting the second period and calculates a distribution pattern indicating a change in the amount of activity over time in the second period, a first calculation unit that obtains a plurality of first distribution patterns that are the plurality of distribution patterns corresponding to each of the plurality of second periods and a second distribution pattern that is the distribution pattern that corresponds to the current second period; ,
a classification unit that classifies the plurality of first distribution patterns into a plurality of groups based on the similarity between the plurality of first distribution patterns;
For each of the plurality of groups, a typical pattern that is a typical distribution pattern of each group is generated based on at least one of the plurality of first distribution patterns belonging to each group, and for each of the plurality of groups, , a pattern in which the amount of activity in each first period in the typical pattern is increased or decreased in accordance with external factor information regarding an external factor that is a factor brought about to the user by matters other than the user is a goal for the user. a second calculation unit that generates a target pattern;
Calculating a difference pattern indicating a difference between the second distribution pattern and a target pattern corresponding to the second distribution pattern among the target patterns calculated for each of the plurality of groups, and applying the difference pattern to the calculated difference pattern. a presentation unit that acquires corresponding advice information for improving the user's lifestyle habits from a storage unit and presents the acquired advice information to the user ;
The external factor information includes at least one of the user's location information, weather forecast information based on the location information, and schedule information regarding the user's schedule.
Habit improvement device. The classification section is
an acquisition unit that acquires a plurality of vectors representing the plurality of first distribution patterns, wherein each of the plurality of vectors is a vector having the activity amount in each of the plurality of first periods as a component; ,
Points in a space spanned by a basis determined from the components correspond to the vector, and a clustering unit clusters and classifies the plurality of points into the plurality of groups based on the plurality of points in the space. , comprising;
The clustering unit includes:
Compare the distance between a certain point and another point in the space, and group the two points with the smallest distance as belonging to the same group,
Compare the distances between a point belonging to a group and multiple points belonging to other groups, and if the minimum distance among these distances is less than or equal to a threshold, the points that have the minimum distance are grouped together. are grouped as the same group,
The habit improvement device according to claim 1. The second calculation unit sets one of the plurality of first distribution patterns belonging to each group as the typical pattern for each of the plurality of groups.
The habit improving device according to claim 1 or 2. The second calculation unit sets, for each of the plurality of groups, a distribution pattern corresponding to a point located at the center or center of gravity of the plurality of points belonging to each group as the typical pattern.
The habit improvement device according to claim 2. The classification unit classifies the plurality of first distribution patterns into the groups based on whether the amount of activity in a specific time period of each first distribution pattern is equal to or higher than a threshold value.
The habit improvement device according to claim 1. The second calculation unit refers to schedule information regarding the user's schedule and location information of the user, determines whether the user deviates from a location scheduled in the schedule, and determines whether the user deviates from the location scheduled in the schedule. If it is determined that there is a facility located at the location indicated by the location information, the target pattern is changed according to the facility located at the location indicated by the location information;
The presentation unit recalculates the difference pattern using the changed target pattern, reacquires advice information corresponding to the recalculated difference pattern from the storage unit, and provides the user with the advice information again . Presenting the obtained advice information,
The habit improving device according to any one of claims 1 to 5. The presentation unit includes:
a calculation unit that calculates the difference pattern by calculating a difference between the second distribution pattern and the target pattern corresponding to the second distribution pattern for each first period;
The habit improvement according to any one of claims 1 to 6, further comprising: an advice generation unit that generates the advice information urging the user to increase the amount of activity when the difference pattern indicates a deficiency in the amount of activity. Device. The first calculation unit uses at least one of the amount of energy consumed by the user and the number of steps taken by the user as the amount of activity.
The habit improving device according to any one of claims 1 to 7. The presentation unit presents the second distribution pattern and the target pattern corresponding to the second distribution pattern to the user as the advice information in a comparable state.
The habit improving device according to any one of claims 1 to 8 . A computer-implemented habit improvement method, the method comprising:
comprising acquiring the amount of activity of the user in each of a plurality of first periods constituting the second period, and calculating a distribution pattern indicating a change in the amount of activity over time in the second period;
Calculating the distribution pattern includes:
Obtaining a plurality of first distribution patterns that are a plurality of the distribution patterns corresponding to each of the plurality of past second periods;
obtaining a second distribution pattern that is the distribution pattern corresponding to the current second period;
including;
classifying the plurality of first distribution patterns into a plurality of groups based on the similarity between the plurality of first distribution patterns;
for each of the plurality of groups, generating a typical pattern that is a typical distribution pattern of each group based on at least one of the plurality of first distribution patterns belonging to each group;
For each of the plurality of groups, a pattern in which the amount of activity in each first period in the typical pattern is increased or decreased in accordance with external factor information regarding an external factor that is a factor brought about to the user by matters other than the user. is generated as a target pattern that is a goal for the user,
Calculating a difference pattern indicating a difference between the second distribution pattern and a target pattern corresponding to the second distribution pattern among the target patterns calculated for each of the plurality of groups, and applying the difference pattern to the calculated difference pattern. acquiring corresponding advice information for improving the user's lifestyle habits from a storage unit, and presenting the acquired advice information to the user ;
The external factor information includes at least one of the user's location information, weather forecast information based on the location information, and schedule information regarding the user's schedule.
How to improve your habits. Classifying the plurality of first distribution patterns into the plurality of groups includes:
obtaining a plurality of vectors representing the plurality of first distribution patterns, each of the plurality of vectors having the activity amount in each of the plurality of first periods as a component;
Points in a space spanned by a basis determined from the components correspond to the vector, and the plurality of points are clustered and classified into the plurality of groups based on the positions of the plurality of points in the space. and,
Clustering and classifying the plurality of points into the plurality of groups includes:
comparing the distance between a certain point and another point in the space and grouping the two points with the smallest distance as belonging to the same group;
Compare the distances between a point belonging to a group and multiple points belonging to other groups, and if the minimum distance among these distances is less than or equal to a threshold, the points that have the minimum distance are grouped together. are grouped as the same group, and
Equipped with
The habit improvement method according to claim 10 . A program for causing a computer to function as each part of the habit improvement device according to any one of claims 1 to 9 .